Data monitoring system and method for wine brewing fermentation processing

Through the analysis of index characteristic sequences during wine fermentation, clustering centers and tolerance are calculated, and the adaptive standard score threshold is adjusted, which solves the problem of low monitoring accuracy of traditional wine fermentation data and improves the quality of wine fermentation.

CN119025950AInactive Publication Date: 2024-11-26SUZHOU UNIV OF SCI & TECH
View PDF 0 Cites 4 Cited by

Patent Information

Application Number
CN202411110952.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-14
Publication Date
2024-11-26
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The accuracy of data monitoring during traditional wine fermentation is low, resulting in poor quality of wine fermentation, mainly because the fixed threshold detection cannot effectively deal with the data floating differences in different fermentation stages.

Method used

By obtaining the index characteristic sequence during the wine fermentation process, calculating the change node and the initial clustering center, performing cluster partitioning, calculating inclusion, and adjusting the adaptive standard score threshold for abnormal monitoring.

Benefits of technology

The monitoring accuracy and product quality of the fermentation process are improved, and the complexity of the data change trends at different fermentation stages is adapted to the adaptive standard score threshold adjustment, which enhances the monitoring sensitivity and accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119025950A_ABST
    Figure CN119025950A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of data processing, in particular to a data monitoring system and method for wine brewing fermentation processing. Obtaining an index change node according to data change features in the index feature sequence; obtaining an initial clustering center according to the position features of the index change nodes; performing clustering according to the initial clustering center to obtain different fermentation stage partitions; and according to the difference characteristics of the change trends of the different index characteristic sequences in the fermentation stage subareas and the distribution characteristics of the different index characteristic sequences, obtaining the comprehensive inclusion degree. The method comprises the following steps: adjusting a preset standard score threshold of different index feature sequences in a fermentation stage partition according to a comprehensive inclusion degree to obtain a self-adaptive standard score threshold; abnormity monitoring of the fermentation process is performed according to the self-adaptive standard score threshold, and the fermentation monitoring accuracy and the quality of the fermentation product are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a data monitoring system and method for wine fermentation processing. Background Art

[0002] Data monitoring during the winemaking fermentation process is crucial to determining brewing quality, improving yield and ensuring product consistency. For example, temperature is crucial to yeast activity and metabolic rate. Monitoring temperature can prevent yeast from overheating or overcooling, which affects the fermentation rate. The pH value during the fermentation process will affect yeast growth and fermentation activity. Monitoring pH value can ensure that the yeast is in a suitable acid-base environment. At the same time, monitoring indicators such as acidity can also reflect the quality of the fermentation process. Therefore, monitoring of various indicators is essential during the winemaking fermentation process.

[0003] Traditionally, data monitoring during the wine fermentation process usually involves collecting and analyzing each indicator separately. For example, the existing standard score method is used to monitor the abnormality of each indicator, and the abnormal situation is judged by analyzing the difference between the standard score value at different times and the set threshold. However, due to the different trends in value changes in different fermentation stages and the differences in data fluctuation, fixed threshold detection can easily lead to low accuracy in indicator monitoring, thus affecting the quality of wine fermentation. Summary of the invention

[0004] In order to solve the technical problem that the accuracy of abnormal monitoring through fixed thresholds in the above-mentioned fermentation index monitoring process is low, which affects the quality of wine fermentation, the purpose of the present invention is to provide a data monitoring system and method for wine fermentation processing, and the technical scheme adopted is as follows:

[0005] Obtain the characteristic sequence of different indicators in the wine fermentation process;

[0006] Obtaining an indicator change node according to the data change characteristics in the indicator feature sequence; obtaining an initial cluster center according to the position characteristics of the indicator change node; performing clustering according to the initial cluster center to obtain different fermentation stage partitions;

[0007] A first tolerance is obtained according to the difference characteristics of the change trends of different indicator feature sequences in the fermentation stage partitions; a second tolerance is obtained according to the distribution characteristics of different indicator feature sequences in the fermentation stage partitions; and a comprehensive tolerance of the fermentation stage partitions is obtained according to the first tolerance and the second tolerance;

[0008] The preset standard score thresholds of different indicator feature sequences in the fermentation stage partitions are adjusted according to the comprehensive inclusiveness to obtain an adaptive standard score threshold; and the fermentation process is monitored for abnormalities according to the adaptive standard score threshold.

[0009] Furthermore, the step of obtaining the indicator change node according to the data change characteristics in the indicator feature sequence includes:

[0010] The first-order derivative function value of each data point in the indicator feature sequence is calculated to obtain a change characteristic value; in the wine fermentation process, the data point whose change characteristic value is closest to the constant 0 in the preset fermentation stage change time period is used as the indicator change node of the indicator feature sequence.

[0011] Furthermore, the step of obtaining the initial cluster center according to the position characteristics of the indicator change node includes:

[0012] Calculate the average value of the horizontal coordinates of all indicator change nodes in any preset fermentation stage change time period to obtain the average horizontal coordinate; calculate the average value of the vertical coordinates of all indicator change nodes in any preset fermentation stage change time period to obtain the average vertical coordinate; construct the initial cluster center of the arbitrary preset fermentation stage change time period based on the average horizontal coordinate and the average vertical coordinate.

[0013] Furthermore, the step of clustering according to the initial cluster center to obtain different fermentation stage partitions includes:

[0014] According to the initial cluster center, the data points of all indicator feature sequences are clustered by the K-means clustering algorithm until the preset clustering termination condition is met to obtain different cluster clusters; the data points in the cluster clusters constitute the fermentation stage partitions.

[0015] Furthermore, the step of obtaining the first tolerance according to the difference characteristics of the change trends of different indicator feature sequences in the fermentation stage partitions includes:

[0016] Calculate the average value of the slope of the data points of the indicator feature sequence in the fermentation stage partition to obtain the average slope feature value; calculate the variance of the average slope feature value of all indicator feature sequences in the fermentation stage partition and normalize them to obtain the change disorder; calculate the slope formed by the first and last data points of the indicator feature sequence in the fermentation stage partition to obtain the change amplitude value; calculate the difference between the maximum and minimum values ​​of the change amplitude value corresponding to the fermentation stage partition and normalize them to obtain the change difference value; calculate the product of the change disorder and a preset first weight to obtain a first coefficient; calculate the product of the change difference value and a preset second weight to obtain a second coefficient; calculate the sum of the first coefficient and the second coefficient to obtain the first inclusiveness of the fermentation stage partition.

[0017] Furthermore, the step of obtaining the second inclusiveness according to the distribution characteristics of different indicator feature sequences in the fermentation stage partitions includes:

[0018] The sum of the number of intersections of all indicator feature sequences in the fermentation stage partition is calculated and normalized to obtain the second inclusiveness of the fermentation stage partition.

[0019] Furthermore, the step of obtaining the comprehensive tolerance of the fermentation stage partition according to the first tolerance and the second tolerance comprises:

[0020] The average value of the first tolerance and the second tolerance is calculated to obtain the comprehensive tolerance of the fermentation stage partition.

[0021] Furthermore, the step of adjusting the preset standard score thresholds of different indicator feature sequences in the fermentation stage partitions according to the comprehensive inclusiveness to obtain the adaptive standard score thresholds includes:

[0022] The product of the comprehensive inclusiveness and the preset standard score threshold is calculated to obtain the adaptive standard score threshold.

[0023] The present invention also proposes a data monitoring system for wine fermentation processing, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement any step of a data monitoring method for wine fermentation processing.

[0024] The present invention has the following beneficial effects:

[0025] In the present invention, obtaining the indicator change node can determine the time point when the indicator undergoes an obvious trend change during the fermentation process, which is conducive to partitioning different fermentation stages; obtaining the initial cluster center can more accurately determine the different fermentation stage partitions, thereby improving the accuracy of the final adaptive standard score threshold setting and preliminarily improving the monitoring accuracy. Obtaining the fermentation stage partition can distinguish the data ranges with different complexity of the indicator feature change trend in the fermentation stage, so that the standard score threshold can be adjusted more accurately. Obtaining the first tolerance can judge the chaotic complexity of the data change trend according to the change trend of different indicator data in the fermentation stage partition; obtaining the second tolerance can judge the chaotic complexity of the data change trend according to the distribution characteristics of the indicator feature sequence in the fermentation stage partition; obtaining the comprehensive tolerance can combine the first tolerance and the second tolerance to more accurately judge the chaotic complexity of the data change trend in the fermentation stage partition, thereby improving the accuracy of the adaptive standard score threshold. Obtaining the adaptive standard score threshold for abnormal monitoring can adaptively adjust the monitoring sensitivity according to the complexity of different data change trends, thereby improving the monitoring accuracy of the fermentation process and the quality of the fermented product. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0027] Figure 1 A flow chart of a data monitoring method for wine fermentation process provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of a data monitoring system and method for wine fermentation treatment proposed by the present invention, its specific implementation method, structure, characteristics and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0029] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0030] The specific scheme of the data monitoring system and method for wine fermentation treatment provided by the present invention is described in detail below with reference to the accompanying drawings.

[0031] See also Figure 1 , which shows a flow chart of a data monitoring method for wine fermentation processing provided by an embodiment of the present invention, the method comprising the following steps:

[0032] Step S1, obtaining the indicator feature sequences of different indicators in the wine fermentation process.

[0033] In the embodiment of the present invention, the implementation scenario is to monitor the indicators in the wine fermentation process to improve the accuracy of the wine fermentation process monitoring and the fermentation quality. First, the indicator feature sequence of different indicators in the wine fermentation process is obtained. In the embodiment of the present invention, the temperature, pH and acidity indicators in the fermentation process are collected, and the collection frequency of different indicators is the same; the indicator data collected in the fermentation process is normalized to avoid excessive differences in the data size of different indicators, which affects the accuracy of the analysis. After normalization, the indicator feature sequence of different indicators is obtained; the implementer can determine the indicators and collection frequency in the wine fermentation process according to the implementation scenario.

[0034] Existing data monitoring usually uses the standard score method to judge abnormalities. The difference between the data point and the average value is calculated and then divided by the standard deviation to obtain the result as the standard score. The more the standard score deviates from 0, the more the data point deviates from the average level. Therefore, the degree of abnormality of the data point can be judged by setting a threshold. However, the numerical changes of indicators at different fermentation stages in the winemaking process are different. Using a fixed standard score threshold is likely to reduce the monitoring accuracy. If the standard score threshold range is large, it may lead to a low sensitivity of abnormal monitoring in the stage of slow data changes, affecting the accuracy of abnormal monitoring; if the standard score threshold range is small, it may lead to a high sensitivity of abnormal monitoring in the stage of drastic data changes, which also affects the accuracy of abnormal monitoring; ultimately leading to poor fermentation quality. Therefore, it is necessary to improve the threshold monitoring using the standard score method to improve the accuracy of abnormal monitoring.

[0035] Step S2, obtaining an indicator change node according to the data change characteristics in the indicator feature sequence; obtaining an initial cluster center according to the position characteristics of the indicator change node; and performing clustering according to the initial cluster center to obtain different fermentation stage partitions.

[0036] The wine fermentation process is divided into multiple stages. Different indicators have different changing trends in different fermentation stages. The conventional wine fermentation process has the initial stage, the middle stage, the high fermentation stage and the late stage. As for the temperature during the fermentation process, the initial stage of wine fermentation is the temperature cooling stage, and the wine needs to be quickly cooled to a suitable fermentation temperature range; in the middle stage of fermentation, due to the active yeast and the heat in the fermentation process, the temperature will fluctuate and rise slightly; in the high fermentation stage, due to the intense activity of the yeast and the heat generated in the wine, the temperature will rise further; in the late stage, as the fermentation approaches the end, the temperature will gradually decrease. The pH value will decrease slightly before adding yeast in the early stage of fermentation; the pH value of the wine is relatively stable in the initial stage; in the high fermentation stage, due to the gradual release of carbon dioxide into the gas phase, reducing the concentration of carbonic acid will cause the pH value to rise, and in the late stage, the pH value will gradually decrease. In the early stage of fermentation, the yeast is not active and does not produce a large amount of acidic metabolites, so the acidity of the wine is low; as the fermentation progresses, the acidity will gradually increase as the yeast releases organic acids in the middle and high fermentation stages; in the late fermentation stage, as the organic acids degrade, the acidity will decrease. Since the indicator change trends in different fermentation stages are different, in order to improve the accuracy of monitoring, it is necessary to first determine the different fermentation stages, so the indicator change nodes are obtained according to the data change characteristics in the indicator feature sequence.

[0037] Preferably, in an embodiment of the present invention, the step of obtaining the index change node includes: calculating the first-order derivative function value of each data point in the index feature sequence to obtain the change characteristic value; the first-order derivative function value can reflect the change rate of the index feature sequence. When the change characteristic value is a constant of 0, it means that the data point is likely to be an inflection point. The data change trend of the index feature sequence at this moment has changed significantly, and the more it conforms to the data change characteristics of the segmentation time point of two adjacent fermentation stages. In the preset fermentation stage change time period during the wine fermentation process, the data point with the change characteristic value closest to the constant 0 is used as the index change node of the index feature sequence; it should be noted that the preset fermentation stage change time period is composed of the last period of a fermentation stage and the initial period of the next fermentation stage. Since the time point of the fermentation stage change cannot be accurately determined, only a time range can be determined according to those skilled in the art. The implementer can determine the preset fermentation stage change time period according to the implementation scenario, and determine the time node when the specific index data changes from the preset fermentation stage change time period; the index change node represents the data point where the data change trend of the index changes significantly, and each indicator has an index change node in the middle of two adjacent fermentation stages.

[0038] Further, according to the analysis of the changing trends of each indicator in the fermentation process, since the indicator change node is the dividing point of two adjacent fermentation stages, there is a significant difference in the changing trends of the indicator data before and after the dividing point. The data characteristics at the indicator change node are the most complex and the data difference is large. The range of the standard score threshold set in the neighborhood of the indicator change node should be larger, so as to avoid the threshold range being too small, resulting in an increased possibility of abnormal monitoring within the range; and for the moment in each fermentation stage, its indicator data is relatively stable, and a smaller standard score threshold range can be set, so as to avoid a large threshold range, resulting in low sensitivity of abnormal monitoring. At the same time, if the data change trends of different indicators in the same period are very different, it means that the fermentation process within the period is more intense and complex, and the stability of each indicator is relatively small, thus requiring a larger standard score threshold range. Therefore, according to the data change trends of different time periods in the fermentation stage, the data ranges with more complex change trends and more stable change trends are segmented, so the initial cluster center is obtained according to the position characteristics of the indicator change node.

[0039] Preferably, in an embodiment of the present invention, the step of obtaining the initial cluster center includes: calculating the average value of the horizontal coordinates of all indicator change nodes in any preset fermentation stage change time period to obtain the average horizontal coordinate; calculating the average value of the vertical coordinates of all indicator change nodes in any preset fermentation stage change time period to obtain the average vertical coordinate; constructing the initial cluster center of any preset fermentation stage change time period according to the average horizontal coordinate and the average vertical coordinate. The horizontal axis of the indicator feature sequence in the two-dimensional plane rectangular coordinate system is different acquisition moments, and the vertical axis is the normalized indicator value. Since there are certain differences in the change trends and reaction times of different indicators, the indicator change nodes of each indicator in the same preset fermentation stage change time period are different, so it is necessary to combine the indicator change nodes of all indicators to determine a comprehensive change node, which is the initial cluster center. The initial cluster center can characterize the central point where the data changes of all indicators in the same preset fermentation stage change time period are more complex; then all indicator feature sequences can be clustered according to the initial cluster center, and data with different change trends can be partitioned, so different fermentation stage partitions are obtained by clustering according to the initial cluster center.

[0040] Further, in an embodiment of the present invention, the step of obtaining the fermentation stage partition includes: clustering the data points of all indicator feature sequences according to the initial cluster center by the K-means clustering algorithm until the preset clustering termination condition is met to obtain different cluster clusters; it should be noted that the K-means belongs to the prior art, and the specific clustering process will not be repeated. Obtaining the initial cluster center for clustering can effectively cluster the data with different change trends, thereby avoiding clustering the data in each fermentation stage into the same cluster cluster and the occurrence of local optimal solutions. In an embodiment of the present invention, the preset clustering termination condition is that less than 50 data points are reallocated to different cluster clusters during the iteration process, thereby avoiding a large change in the cluster center during the iterative clustering process, affecting the purpose of clustering data with different change trends. The implementer can determine the preset clustering termination condition according to the implementation scenario. After clustering is completed, the data points in each cluster cluster form a fermentation stage partition, and the data change trends of the indicator feature sequences of the same indicator in different fermentation stage partitions are quite different.

[0041] Step S3, obtaining a first tolerance according to the difference characteristics of the change trends of different indicator feature sequences in the fermentation stage partitions; obtaining a second tolerance according to the distribution characteristics of different indicator feature sequences in the fermentation stage partitions; and obtaining a comprehensive tolerance of the fermentation stage partitions according to the first tolerance and the second tolerance.

[0042] Due to the differences in the changing trends of the characteristic sequences of indicators in different fermentation stage partitions, the more complex the changing trends are, the larger the standard score threshold should be set to avoid excessive sensitivity of abnormal monitoring. When there are obvious differences in the changing trends of the characteristic sequences of indicators of different indicators in the same fermentation stage partition, it means that the fermentation response in the fermentation stage partition is more intense, the changing trends of different indicators are more complex, and the data confusion is greater. Therefore, the first tolerance is obtained based on the difference characteristics of the changing trends of the characteristic sequences of different indicators in the fermentation stage partition.

[0043] Preferably, in an embodiment of the present invention, the step of obtaining the first inclusiveness includes: calculating the average value of the slope of the data points of the indicator feature sequence in the fermentation stage partition to obtain the average slope characteristic value; it should be noted that the data point slope is calculated by the existing point slope method, and the slope of the data point is obtained according to the data point and the data point at the previous moment; the average slope characteristic value reflects the average change characteristics of the indicator feature sequence in the partial data of any fermentation stage partition. Calculate the variance of the average slope characteristic values ​​of all indicator feature sequences in the fermentation stage partition and normalize them to obtain the change disorder degree; when the difference in the average slope characteristic values ​​of different indicators is greater, the change disorder degree is closer to 1, which means that the change trend of different data in the fermentation stage partition is more chaotic, and the change trend of the fermentation reflection period corresponding to the fermentation stage partition is more complex. Calculate the slope formed by the first and last two data points of the indicator feature sequence in the fermentation stage partition to obtain the change amplitude value; the change amplitude value reflects the overall change trend of the indicator in the fermentation stage partition. Calculate the difference between the maximum and minimum values ​​of the corresponding change amplitude values ​​in the fermentation stage partition and normalize them to obtain the change difference value; the larger the change difference value, the greater the difference in the change trends of different indicators within the fermentation stage partition, and the more complex the change trends of different data in the fermentation stage partition. Calculate the product of the change confusion and the preset first weight to obtain the first coefficient; calculate the product of the change difference value and the preset second weight to obtain the second coefficient; calculate the sum of the first coefficient and the second coefficient to obtain the first tolerance of the fermentation stage partition; in an embodiment of the present invention, the preset first weight is 0.6, and the preset second weight is 0.4. The implementer can determine it according to the implementation scenario. The larger the first tolerance, the more chaotic and complex the data change trends of different indicators in the fermentation stage partition, and the more intense the fermentation process in this period, so a larger standard score threshold range is needed to avoid high sensitivity and monitor the indicator data in this period as an abnormal state. The formula for obtaining the first tolerance includes:

[0044] D=a1*norm(S)+a2*norm(HM)

[0045] Wherein, D represents the first inclusiveness of the fermentation stage partition, a1 represents the preset first weight, a2 represents the preset second weight, norm() represents the normalization function, S represents the variance of the average slope characteristic value of all indicator characteristic sequences in the fermentation stage partition, norm(S) represents the degree of change confusion, a1*norm(S) represents the first coefficient; H represents the maximum value of the change amplitude value, M represents the minimum value of the change amplitude value, norm(HM) represents the change difference value, and a2*norm(HM) represents the second coefficient.

[0046] Furthermore, when there are more data points where the indicator feature sequences of different indicators in the fermentation stage partition intersect, it means that the data changes of different indicator feature sequences in the fermentation stage partition are more complex and the trends are more chaotic; therefore, the second tolerance is obtained according to the distribution characteristics of different indicator feature sequences in the fermentation stage partition. Preferably, in an embodiment of the present invention, the step of obtaining the second tolerance includes: calculating the sum of the number of intersections of all indicator feature sequences in the fermentation stage partition and normalizing them to obtain the second tolerance of the fermentation stage partition. The intersection refers to the data point where any two indicator feature sequences in the fermentation stage partition intersect. The more the number of intersections, the greater the second tolerance, which means that the data change trend in the fermentation stage partition is more complex, and a larger standard score threshold range is required to avoid high sensitivity.

[0047] Then, the comprehensive inclusiveness of the fermentation stage partition can be obtained according to the first inclusiveness and the second inclusiveness. Preferably, in the embodiment of the present invention, the step of obtaining the comprehensive inclusiveness includes: calculating the average of the first inclusiveness and the second inclusiveness to obtain the comprehensive inclusiveness of the fermentation stage partition. When the comprehensive inclusiveness is smaller, it means that the change trend of the characteristic sequence of the indicator in the fermentation stage partition is more stable and the change trend is less complex, so a smaller standard score threshold range can be used to ensure monitoring sensitivity; when the comprehensive inclusiveness is larger, it means that the change trend of the characteristic sequence of the indicator in the fermentation stage partition is more complex and the data change amplitude is more obvious, so a larger standard score threshold range can be used to avoid excessive monitoring sensitivity.

[0048] Step S4, adjusting the preset standard score thresholds of different indicator feature sequences in the fermentation stage partitions according to the comprehensive inclusiveness to obtain an adaptive standard score threshold; and performing abnormal monitoring of the fermentation process according to the adaptive standard score threshold.

[0049] It can be seen from the analysis of step S3 that the comprehensive tolerance and the range of the standard score threshold are positively correlated, so the preset standard score threshold of different indicator feature sequences in the fermentation stage partition is adjusted according to the comprehensive tolerance to obtain an adaptive standard score threshold. Preferably, in an embodiment of the present invention, the step of obtaining the adaptive standard score threshold includes: calculating the product of the comprehensive tolerance and the preset standard score threshold to obtain the adaptive standard score threshold. When the comprehensive tolerance is larger, the range of the adaptive standard score threshold is larger, so the monitoring sensitivity is lower, and it is avoided that it is misjudged as an abnormal state during the complex stage of data change; when the comprehensive tolerance is smaller, the range of the adaptive standard score threshold is smaller, so the monitoring sensitivity is higher, and the monitoring sensitivity in the period when the data change trend is relatively stable is improved to avoid abnormal situations not being monitored. The implementer can determine the preset standard score threshold according to the implementation scenario, which is not limited here. So far, an indicator feature sequence corresponds to different adaptive standard score thresholds in different fermentation stage partitions, and then the abnormal monitoring of the fermentation process is performed according to the adaptive standard score threshold. When the calculated standard score exceeds the adaptive standard score threshold at the corresponding moment, it can be judged as an abnormal state, and the abnormal state at that moment is improved in the subsequent batches of wine fermentation process. The abnormality monitoring of the fermentation process is carried out through adaptive standard score threshold, which improves the monitoring accuracy and the quality of the fermentation product.

[0050] In summary, the embodiment of the present invention provides a data monitoring method for wine fermentation processing; obtains an indicator change node according to the data change characteristics in the indicator feature sequence; obtains an initial cluster center according to the position characteristics of the indicator change node; clusters according to the initial cluster center to obtain different fermentation stage partitions; obtains a comprehensive tolerance according to the difference characteristics of the change trends of different indicator feature sequences in the fermentation stage partitions and the distribution characteristics of different indicator feature sequences. The present invention adjusts the preset standard score thresholds of different indicator feature sequences in the fermentation stage partitions according to the comprehensive tolerance to obtain an adaptive standard score threshold; performs abnormal monitoring of the fermentation process according to the adaptive standard score threshold, thereby improving the accuracy of fermentation monitoring and the quality of fermented products.

[0051] The present invention also proposes a data monitoring system for wine fermentation treatment, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement any step of a data monitoring method for wine fermentation treatment.

[0052] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0053] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A data monitoring method for wine fermentation process, characterized in that: The method comprises the following steps: Obtain the characteristic sequence of different indicators in the wine fermentation process; Obtaining an indicator change node according to the data change characteristics in the indicator feature sequence; obtaining an initial cluster center according to the position characteristics of the indicator change node; performing clustering according to the initial cluster center to obtain different fermentation stage partitions; A first tolerance is obtained according to the difference characteristics of the change trends of different indicator feature sequences in the fermentation stage partitions; a second tolerance is obtained according to the distribution characteristics of different indicator feature sequences in the fermentation stage partitions; and a comprehensive tolerance of the fermentation stage partitions is obtained according to the first tolerance and the second tolerance; The preset standard score thresholds of different indicator feature sequences in the fermentation stage partitions are adjusted according to the comprehensive inclusiveness to obtain an adaptive standard score threshold; and the fermentation process is monitored for abnormalities according to the adaptive standard score threshold.

2. The data monitoring method for wine fermentation process according to claim 1 is characterized in that: The step of obtaining an indicator change node according to the data change characteristics in the indicator feature sequence comprises: The first-order derivative function value of each data point in the indicator feature sequence is calculated to obtain a change characteristic value; in the wine fermentation process, the data point whose change characteristic value is closest to the constant 0 in the preset fermentation stage change time period is used as the indicator change node of the indicator feature sequence.

3. The data monitoring method for wine fermentation process according to claim 2 is characterized in that: The step of obtaining the initial cluster center according to the position characteristics of the indicator change node comprises: Calculate the average value of the horizontal coordinates of all indicator change nodes in any preset fermentation stage change time period to obtain the average horizontal coordinate; calculate the average value of the vertical coordinates of all indicator change nodes in any preset fermentation stage change time period to obtain the average vertical coordinate; construct the initial cluster center of the arbitrary preset fermentation stage change time period based on the average horizontal coordinate and the average vertical coordinate.

4. The data monitoring method for wine fermentation process according to claim 1 is characterized in that: The step of clustering according to the initial cluster center to obtain different fermentation stage partitions comprises: According to the initial cluster center, the data points of all indicator feature sequences are clustered by the K-means clustering algorithm until the preset clustering termination condition is met to obtain different cluster clusters; the data points in the cluster clusters constitute the fermentation stage partitions.

5. The data monitoring method for wine fermentation process according to claim 1 is characterized in that: The step of obtaining the first tolerance according to the difference characteristics of the change trends of different indicator feature sequences in the fermentation stage partitions comprises: Calculate the average value of the slope of the data points of the indicator feature sequence in the fermentation stage partition to obtain the average slope feature value; calculate the variance of the average slope feature value of all indicator feature sequences in the fermentation stage partition and normalize them to obtain the change disorder; calculate the slope formed by the first and last data points of the indicator feature sequence in the fermentation stage partition to obtain the change amplitude value; calculate the difference between the maximum and minimum values ​​of the change amplitude value corresponding to the fermentation stage partition and normalize them to obtain the change difference value; calculate the product of the change disorder and a preset first weight to obtain a first coefficient; calculate the product of the change difference value and a preset second weight to obtain a second coefficient; calculate the sum of the first coefficient and the second coefficient to obtain the first inclusiveness of the fermentation stage partition.

6. The data monitoring method for wine fermentation process according to claim 1 is characterized in that: The step of obtaining the second tolerance according to the distribution characteristics of different indicator feature sequences in the fermentation stage partitions comprises: The sum of the number of intersections of all indicator feature sequences in the fermentation stage partition is calculated and normalized to obtain the second inclusiveness of the fermentation stage partition.

7. The data monitoring method for wine fermentation process according to claim 1 is characterized in that: The step of obtaining the comprehensive tolerance of the fermentation stage partition according to the first tolerance and the second tolerance comprises: The average value of the first tolerance and the second tolerance is calculated to obtain the comprehensive tolerance of the fermentation stage partition.

8. The data monitoring method for wine fermentation process according to claim 1 is characterized in that: The step of adjusting the preset standard score thresholds of different indicator feature sequences in the fermentation stage partitions according to the comprehensive inclusiveness to obtain the adaptive standard score thresholds comprises: The product of the comprehensive inclusiveness and the preset standard score threshold is calculated to obtain the adaptive standard score threshold.

9. A data monitoring system for wine fermentation processing, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 8.

Citation Information

Cited By

  • Microbial fermentation temperature high-precision monitoring method and system

    CN120760887A

  • Fermentation process control method based on real-time monitoring of physicochemical property change of dough

    CN120871801A

  • Method for controlling fermentation process based on real-time monitoring of changes in dough physicochemical properties

    CN120871801B

  • Production line scheduling optimization method based on artificial intelligence

    CN121169006A